Method and device for calibrating extrinsic parameters of vehicle-mounted surround-view cameras, vehicle and storage medium

Abstract

Embodiments of the present application disclose a vehicle-mounted surround-view camera extrinsic parameter calibration method and device, a vehicle and a storage medium. The method comprises: determining the extrinsic parameters of a first camera and the extrinsic parameters of a second camera; determining a first pose initial value, a second pose initial value, a third pose initial value and a fourth pose initial value; and determining the extrinsic parameters of a third camera and the extrinsic parameters of a fourth camera according to the extrinsic parameters of the first camera, the extrinsic parameters of the second camera, the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value. By determining the extrinsic parameters of the first camera and the second camera according to the visual odometry information of the first camera and the second camera first, and then determining the extrinsic parameters of the third camera and the fourth camera, the problem of large computing power and long time consumption caused by the visual odometry calculation of the four cameras to determine the extrinsic parameters can be avoided, and the real-time performance of the extrinsic parameter calibration is improved. After the initial values of the extrinsic parameters of the four cameras are calculated, the initial values of the extrinsic parameters of the four cameras are jointly optimized, and the extrinsic parameter calibration accuracy is improved.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of automatic driving, in particular to a kind of vehicle-mounted ring view camera's external parameter calibration method, device, vehicle and storage medium. BACKGROUND

[0002] In intelligent driving assistance system, the calibration of vehicle-mounted ring view camera is the basis of most intelligent perception functions, so the stability and precision of ring view camera calibration directly affect the precision of visual perception function. Camera calibration can be divided into intrinsic parameter calibration and extrinsic parameter calibration. Camera intrinsic parameter is the inherent parameter of camera. Camera extrinsic parameter refers to the parameter representing the pose relationship of camera coordinate system relative to vehicle body coordinate system. Vehicle body coordinate system refers to the coordinate system established by taking the vehicle itself as the reference system, which can also be regarded as the coordinate system used to describe the relative pose relationship between objects around the vehicle and the vehicle. Camera coordinate system refers to a three-dimensional rectangular coordinate system with the focusing center of the camera as the origin and the optical axis as the Z axis.

[0003] At present, the extrinsic parameter calibration method of vehicle-mounted ring view camera can be divided into three types: first, a camera calibration method based on a specific site. This method basically uses a specific calibration template placed in the scene to calibrate by extracting feature points of the calibration template. However, this method has high site construction cost and can only be applied to a vehicle stationary environment. Second, a camera calibration method based on vanishing point. This method calculates the vanishing point in the image using parallel information in the environment, and then calculates the extrinsic parameter of the camera based on the vanishing point. However, this method relies heavily on special signs with parallel information, such as lane lines, and has limited application scenarios. Third, a camera calibration method based on scene map fusion. This method calculates the relative pose between cameras through map information fusion, and each camera needs to run a odometer to establish an environmental map. This method requires high computing power and takes a long time, which is not conducive to real-time online calibration.

[0004] Therefore, how to improve the real-time and universality of the extrinsic parameter calibration of vehicle-mounted ring view camera is a technical problem to be solved. SUMMARY

[0005] Embodiments of the present application provide a kind of vehicle-mounted ring view camera's external parameter calibration method, device, vehicle and storage medium, to the real-time and universality of vehicle-mounted ring view camera extrinsic parameter calibration.

[0006] According to an aspect of embodiments of the present application, a vehicle-mounted ring view camera extrinsic parameter calibration method is provided, applied to a vehicle, the vehicle comprising a vehicle-mounted ring view camera, the vehicle-mounted ring view camera comprising a first camera, a second camera, a third camera and a fourth camera, the first camera position and the third camera position, the fourth camera position are adjacent, the second camera position and the third camera position, the fourth camera position are adjacent, and the method comprises:

[0007] determine the extrinsic parameters of the first camera and the extrinsic parameters of the second camera according to the vehicle body odometry information, the first visual odometry information of the first camera, and the second visual odometry information of the second camera;

[0008] determine the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value according to the vehicle body odometry information, the first visual odometry information, the second visual odometry information, the extrinsic parameters of the first camera, and the extrinsic parameters of the second camera;

[0009] determine the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value according to the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value;

[0010] determine the extrinsic parameters of the third camera and the extrinsic parameters of the fourth camera according to the extrinsic parameters of the first camera, the extrinsic parameters of the second camera, the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value.

[0011] According to another aspect of the embodiment of the present application, there is provided a device for calibrating extrinsic parameters of a vehicle-mounted surround-view camera, configured in a vehicle, wherein the vehicle comprises a vehicle-mounted surround-view camera, and the vehicle-mounted surround-view camera comprises a first camera, a second camera, a third camera, and a fourth camera, the first camera is located adjacent to the third camera and the fourth camera, and the second camera is located adjacent to the third camera and the fourth camera, and the device comprises:

[0012] a first extrinsic parameter determination module, configured to determine the extrinsic parameters of the first camera and the extrinsic parameters of the second camera according to the vehicle body odometry information, the first visual odometry information of the first camera, and the second visual odometry information of the second camera;

[0013] a pose determination module, configured to determine the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value according to the vehicle body odometry information, the first visual odometry information, the second visual odometry information, the extrinsic parameters of the first camera, and the extrinsic parameters of the second camera;

[0014] an initial value determination module, configured to determine the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value according to the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value;

[0015] The second extrinsic parameter determination module is configured to determine the extrinsic parameters of the third camera and the fourth camera according to the extrinsic parameter of the first camera, the extrinsic parameter of the second camera, the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value.

[0016] According to another aspect of embodiments of the present application, a vehicle is provided, the vehicle comprising:

[0017] a vehicle-mounted surround-view camera;

[0018] at least one processor; and

[0019] a memory connected to the at least one processor; wherein

[0020] the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform the method for calibrating the extrinsic parameters of the vehicle-mounted surround-view camera according to any one of the embodiments of the present application.

[0021] According to another aspect of embodiments of the present application, a computer readable storage medium is provided, the computer readable storage medium stores computer instructions for enabling a processor to implement the method for calibrating the extrinsic parameters of the vehicle-mounted surround-view camera according to any one of the embodiments of the present application when the processor executes the computer instructions.

[0022] The technical scheme of the embodiment of the application determines the extrinsic parameters of the first camera and the extrinsic parameters of the second camera according to the vehicle body odometer information, the first visual odometer information of the first camera, and the second visual odometer information of the second camera; determines the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value according to the vehicle body odometer information, the first visual odometer information, the second visual odometer information, the extrinsic parameters of the first camera, and the extrinsic parameters of the second camera; determines the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value according to the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value; and determines the extrinsic parameters of the third camera and the extrinsic parameters of the fourth camera according to the extrinsic parameters of the first camera, the extrinsic parameters of the second camera, the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value. The technical scheme can avoid the problems of large computing power and long time consumption caused by the visual odometer calculation of the four cameras to determine the extrinsic parameters, thereby improving the real-time performance of the extrinsic parameter calibration. In addition, the scheme can be applied to the extrinsic parameter calibration in any scene, can avoid the problem of relying on road-specific markers, and improves the universality of the extrinsic parameter calibration.

[0023] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the application, nor is it used to limit the scope of the application. Other features of the application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0025] Figure 1 A flowchart of a vehicle-mounted surround view camera extrinsic parameter calibration method provided by the first embodiment of the application;

[0026] Figure 2 An implementation schematic diagram of a closed loop relationship construction provided by the first embodiment of the application;

[0027] Figure 3 An implementation schematic diagram of a four-camera provided by the first embodiment of the application;

[0028] Figure 4This is a flowchart illustrating a method for calibrating the extrinsic parameters of a vehicle-mounted surround-view camera according to Embodiment 2 of the present invention.

[0029] Figure 5 This is a schematic diagram of the external parameter calibration device for a vehicle-mounted surround view camera provided in Embodiment 3 of the present invention;

[0030] Figure 6 This is a structural schematic diagram of a vehicle provided in Embodiment 4 of the present invention. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0033] Example 1

[0034] Figure 1 This is a flowchart illustrating an extrinsic parameter calibration method for a vehicle-mounted surround-view camera according to Embodiment 1 of the present invention. This method is applicable to calibrating the extrinsic parameters of a vehicle-mounted surround-view camera. The method can be executed by an extrinsic parameter calibration device for the vehicle-mounted surround-view camera, which can be implemented by software and / or hardware and is generally integrated into the vehicle. The vehicle includes a vehicle-mounted surround-view camera, which includes a first camera, a second camera, a third camera, and a fourth camera. The positions of the first camera, the third camera, and the fourth camera are adjacent to each other, and the positions of the second camera, the third camera, and the fourth camera are also adjacent to each other. The positions can be understood as the locations where the cameras are situated.

[0035] like Figure 1 As shown in Embodiment 1 of the present invention, a method for calibrating the extrinsic parameters of a vehicle-mounted surround-view camera is provided. The method includes the following steps:

[0036] In S110, the extrinsic parameters of the first camera and the extrinsic parameters of the second camera are determined according to the vehicle body odometry information, the first visual odometry information of the first camera, and the second visual odometry information of the second camera.

[0037] In the embodiment, the first camera, the second camera, the third camera, and the fourth camera can be understood as cameras for image acquisition in the front view, the rear view, the left view, and the right view of the vehicle body. The setting of the direction corresponding to each camera is not limited, but the positions of the first camera, the third camera, and the fourth camera are adjacent, and the positions of the second camera, the third camera, and the fourth camera are adjacent. That is, if the first camera is a front view camera, the third camera can be a left view camera or a right view camera adjacent to the position of the front view camera. Correspondingly, if the second camera is a rear view camera, the fourth camera can be a left view camera or a right view camera adjacent to the second camera. It can be understood that the positions corresponding to each camera are different, for example, two cameras cannot be left view cameras.

[0038] The first visual odometry information can be understood as the visual odometry information corresponding to the first camera. The second visual odometry information can be understood as the visual odometry information corresponding to the second camera.

[0039] The vehicle body odometry information can be understood as information associated with the vehicle body odometry of the vehicle. The vehicle body odometry can be understood as information representing the change in the moving position of the vehicle based on the vehicle body. The visual odometry information can be understood as information associated with the visual odometry of the corresponding camera. The visual odometry can be understood as information representing the change in the moving position of the camera based on the camera and map points. The specific content of the vehicle body odometry information and the visual odometry information is not limited here. For example, the vehicle body odometry information can include vehicle speed, vehicle body motion trajectory, and the like. The visual odometry information can include map points, camera motion trajectory, and the like. The vehicle body motion trajectory can be understood as information representing the moving trajectory of the vehicle motion. The camera motion trajectory can be understood as information representing the moving trajectory of the camera motion.

[0040] Herein, how to determine the extrinsic parameters of the first camera and the extrinsic parameters of the second camera based on the vehicle body odometry information, the first visual odometry information of the first camera and the second visual odometry information of the second camera is not specifically limited. For example, the vehicle body odometry information can be obtained by a real-time kinematic (RTK) method, an inertial measurement unit (IMU) method, a wheel speed meter method, etc. Then, the first visual odometry information and the second visual odometry information can be obtained by corresponding visual odometry calculation methods based on images collected by the first camera and the second camera. Finally, the extrinsic parameters of the first camera and the extrinsic parameters of the second camera can be obtained by corresponding extrinsic parameter algorithms (for example, an algorithm based on a pre-set hand-eye calibration principle) based on the vehicle body odometry information, the first visual odometry information and the second visual odometry information.

[0041] In S120, first pose initial value, second pose initial value, third pose initial value and fourth pose initial value are determined based on the vehicle body odometry information, the first visual odometry information, the second visual odometry information, the extrinsic parameters of the first camera and the extrinsic parameters of the second camera.

[0042] In this embodiment, the first pose initial value can be understood as an initial value representing the relative pose between the first camera and the third camera. The second pose initial value can be understood as an initial value representing the relative pose between the first camera and the fourth camera. The third pose initial value is an initial value representing the relative pose between the second camera and the third camera. The fourth pose initial value is an initial value representing the relative pose between the second camera and the fourth camera.

[0043] The first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value are determined according to the vehicle body odometer information, the first visual odometer information, the second visual odometer information, the extrinsic parameters of the first camera and the extrinsic parameters of the second camera. For example, the map points in the overlapping area of the collection ranges of the first camera and the third camera, the map points in the overlapping area of the collection ranges of the first camera and the fourth camera, the map points in the overlapping area of the collection ranges of the second camera and the third camera, and the map points in the overlapping area of the collection ranges of the second camera and the fourth camera are determined according to the vehicle body odometer information, the map points in the visual odometer information and the image collection ranges of the cameras. The determined map points are converted into the camera coordinate system according to the extrinsic parameters of the first camera and the second camera, and then converted into the vehicle body coordinate system, so as to obtain the map points in the vehicle body coordinate system. The extrinsic parameters of the third camera and the fourth camera are determined according to the map points in the vehicle body coordinate system and the images collected by the cameras. The first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value are determined according to the extrinsic parameters of the first camera and the second camera and the extrinsic parameters of the third camera and the fourth camera.

[0044] In S130, first target pose initial value, second target pose initial value, third target pose initial value and fourth target pose initial value are determined according to the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value.

[0045] In this embodiment, the target pose initial value can be understood as the pose initial value used to determine the extrinsic parameters of the third camera and the fourth camera. The first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value are not specifically limited, and for example, the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value can be directly determined as the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value, respectively. Alternatively, the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value can be optimized by a corresponding optimization algorithm, so as to determine the optimized pose initial values as the target pose initial values.

[0046] In S140, the extrinsic parameters of the third camera and the fourth camera are determined according to the extrinsic parameters of the first camera, the extrinsic parameters of the second camera, the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value.

[0047] In the embodiment, how to determine the extrinsic parameters of the third camera and the fourth camera according to the extrinsic parameters of the first camera, the extrinsic parameters of the second camera, the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value is not limited. For example, the extrinsic parameters of the third camera can be obtained according to the extrinsic parameters of the first camera and the first target pose initial value, or the extrinsic parameters of the second camera and the fourth target pose initial value, by using a corresponding extrinsic parameter algorithm. Specifically, the first camera, the third camera and the vehicle body can form a closed loop relationship, and the extrinsic parameters of the third camera can be calculated according to the formula wherein represents the extrinsic parameters of the third camera, represents the first target pose initial value, represents the extrinsic parameters of the first camera.

[0048] The extrinsic parameters of the fourth camera can be obtained according to the extrinsic parameters of the first camera and the second target pose initial value, or the extrinsic parameters of the second camera and the fourth target pose initial value, by using a corresponding extrinsic parameter algorithm. Specifically, the second camera, the fourth camera and the vehicle body can form a closed loop relationship, and the extrinsic parameters of the fourth camera can be calculated according to the formula wherein represents the extrinsic parameters of the fourth camera, represents the fourth target pose initial value, represents the extrinsic parameters of the second camera.

[0049] Figure 2 An implementation schematic diagram of a closed loop relationship construction provided by the first embodiment of the present application is shown in FIG. 1. Figure 2 As shown in FIG. 1, taking the front-view camera, the left-view camera and the vehicle body as an example, the front-view camera (i.e. the first camera) is adjacent to the left-view camera (i.e. the third camera), and the front-view camera, the left-view camera and the vehicle body form a closed loop relationship.

[0050] Figure 3 An implementation schematic diagram of a four-way camera provided by the first embodiment of the present application is shown in FIG. 2. Figure 3 As shown in FIG. 2, the extrinsic parameters of the front-view camera are the extrinsic parameters of the rear-view camera are the extrinsic parameters of the left-view camera are the extrinsic parameters of the right-view camera are

[0051] The embodiment one of the present application provides a kind of vehicle-mounted ring view camera's external parameter calibration method, according to the vehicle body odometer information of vehicle, the first visual odometer information of first camera and the second visual odometer information of second camera, the external parameter of first camera and the external parameter of second camera are determined;According to vehicle body odometer information, first visual odometer information, second visual odometer information, the external parameter of first camera and the external parameter of second camera, first pose initial value, second pose initial value, third pose initial value and fourth pose initial value are determined;First target pose initial value, second target pose initial value, third target pose initial value and fourth target pose initial value are determined according to first pose initial value, second pose initial value, third pose initial value and fourth pose initial value;The external parameter of third camera and the external parameter of fourth camera are determined according to the external parameter of first camera, the external parameter of second camera, first target pose initial value, second target pose initial value, third target pose initial value and fourth target pose initial value.This technical solution can avoid the problem of large computing power and long time consumption caused by visual odometer calculation of four cameras to determine external parameter by determining the corresponding external parameter according to the visual odometer information of first camera and second camera first, and then determining the external parameter of third camera and fourth camera according to the external parameter of first camera and second camera, to improve the real-time performance of external parameter calibration.In addition, the present solution can be applied to external parameter calibration in any scene, and can avoid the problem of relying on road specific markers, improving the versatility of external parameter calibration.

[0052] Embodiment two

[0053] Figure 4 The flowchart of the vehicle-mounted ring view camera's external parameter calibration method provided by the embodiment two of the present application is shown, and the embodiment two is refined based on the above-mentioned embodiments.In this embodiment, the process of determining the external parameter of the first camera and the external parameter of the second camera according to the vehicle body odometer information of the vehicle, the first visual odometer information of the first camera and the second visual odometer information of the second camera, and the process of determining the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value according to the vehicle body odometer information, the first visual odometer information, the second visual odometer information, the external parameter of the first camera and the external parameter of the second camera are specifically described.It should be noted that the technical details not described in detail in this embodiment can be referred to any of the above-mentioned embodiments.For example, as shown in the figure, the method comprises: Figure 4

[0054] S210, obtain vehicle body odometer information and images collected by each camera.

[0055] In this embodiment, the vehicle can obtain the vehicle body odometer information of the vehicle through a corresponding algorithm.The vehicle can obtain the images collected by each camera in the vehicle-mounted ring view camera.

[0056] ​S220, determine the first visual odometry information and the second visual odometry information based on the images collected by the first camera and the second camera.

[0057] In this embodiment, the vehicle can obtain the corresponding first visual odometry information through a corresponding algorithm based on the images collected by the first camera; the vehicle can obtain the corresponding second visual odometry information through a corresponding algorithm based on the images collected by the second camera; and the present disclosure does not limit this.

[0058] S230, determine the extrinsic parameters of the first camera and the extrinsic parameters of the second camera by setting a hand-eye calibration algorithm based on the vehicle body odometry information and the determined visual odometry information.

[0059] In this embodiment, setting a hand-eye calibration algorithm can be understood as a pre-set extrinsic parameter calculation method based on the principle of hand-eye calibration. The vehicle can obtain the extrinsic parameters of the first camera by setting a hand-eye calibration algorithm based on the vehicle body odometry information and the first visual odometry information; the vehicle can obtain the extrinsic parameters of the second camera by setting a hand-eye calibration algorithm based on the vehicle body odometry information and the second visual odometry information; and the present disclosure does not limit this.

[0060] S240, determine the first effective map point, the second effective map point, the third effective map point and the fourth effective map point according to the vehicle body odometry information, the first map point in the first visual odometry information, the second map point in the second visual odometry information and the collection range of each camera.

[0061] In this embodiment, a map point can be understood as a three-dimensional point corresponding to a feature point of an image collected by a camera. Correspondingly, the first map point can be understood as a map point collected by the first camera. The second map point can be understood as a map point collected by the second camera.

[0062] The first effective map point is a map point in the overlapping area of the first camera and the third camera. The second effective map point is a map point in the overlapping area of the first camera and the fourth camera. The third effective map point is a map point in the overlapping area of the second camera and the third camera. The fourth effective map point is a map point in the overlapping area of the second camera and the fourth camera.

[0063] The determination of the first effective map point, the second effective map point, the third effective map point and the fourth effective map point is not specifically limited here; for example, a scale factor can be determined according to the vehicle body odometer information and the visual odometer information, the first map point and the second map point are respectively converted based on the scale factor to obtain the first real map point and the second real map point under the real world map corresponding to the scale factor; then, the overlapping area between the first camera and the third camera (i.e., the first overlapping area), the overlapping area between the first camera and the fourth camera (i.e., the second overlapping area), the overlapping area between the second camera and the third camera (i.e., the third overlapping area), and the overlapping area between the second camera and the fourth camera (i.e., the fourth overlapping area) are determined; finally, the map point of the first real map point in the first overlapping area can be determined as the first effective map point, the map point of the first real map point in the second overlapping area can be determined as the second effective map point, the map point of the second real map point in the third overlapping area can be determined as the third effective map point, and the map point of the second real map point in the fourth overlapping area can be determined as the fourth effective map point.

[0064] S250, according to the poses of the first visual odometer information and the second visual odometer information, the first effective map point and the second effective map point are converted to the first camera coordinate system to obtain the first coordinate point and the second coordinate point, and the third effective map point and the fourth effective map point are converted to the second camera coordinate system to obtain the third coordinate point and the fourth coordinate point.

[0065] In this embodiment, the first coordinate point can be understood as the coordinate point obtained by converting the first effective map point to the first camera coordinate system. The second coordinate point can be understood as the coordinate point obtained by converting the second effective map point to the first camera coordinate system. The third coordinate point can be understood as the coordinate point obtained by converting the third effective map point to the second camera coordinate system. The fourth coordinate point can be understood as the coordinate point obtained by converting the fourth effective map point to the second camera coordinate system.

[0066] S260, according to the extrinsic parameters of the first camera and the extrinsic parameters of the second camera, the first coordinate point, the second coordinate point, the third coordinate point and the fourth coordinate point are respectively converted to the vehicle body coordinate system to obtain the first vehicle body coordinate point, the second vehicle body coordinate point, the third vehicle body coordinate point and the fourth vehicle body coordinate point.

[0067] In this embodiment, the first vehicle body coordinate point can be understood as the point obtained by converting the first effective map point to the vehicle body coordinate system. The second vehicle body coordinate point can be understood as the point obtained by converting the second effective map point to the vehicle body coordinate system. The third vehicle body coordinate point can be understood as the point obtained by converting the third effective map point to the vehicle body coordinate system. The fourth vehicle body coordinate point can be understood as the point obtained by converting the fourth effective map point to the vehicle body coordinate system.

[0068] The extrinsic parameter of the first camera represents a pose relationship of the vehicle body coordinate system relative to the first camera coordinate system, and an inverse of the extrinsic parameter of the first camera represents a pose relationship of the first camera coordinate system relative to the vehicle body coordinate system. On this basis, the first coordinate point can be converted into a point in the vehicle body coordinate system, i.e., the first vehicle body coordinate point, and the second coordinate point can be converted into a point in the vehicle body coordinate system, i.e., the second vehicle body coordinate point, based on the pose relationship. Correspondingly, the extrinsic parameter of the second camera represents a pose relationship of the vehicle body coordinate system relative to the second camera coordinate system, and an inverse of the extrinsic parameter of the second camera represents a pose relationship of the second camera coordinate system relative to the vehicle body coordinate system. On this basis, the third coordinate point can be converted into a point in the vehicle body coordinate system, i.e., the third vehicle body coordinate point, and the fourth coordinate point can be converted into a point in the vehicle body coordinate system, i.e., the fourth vehicle body coordinate point, based on the pose relationship.

[0069] S270, determining the first target extrinsic parameter initial value and the second target extrinsic parameter initial value of the third camera and the third target extrinsic parameter initial value and the fourth target extrinsic parameter initial value of the fourth camera according to the images collected by the cameras, the first vehicle body coordinate point, the second vehicle body coordinate point, the third vehicle body coordinate point, and the fourth vehicle body coordinate point.

[0070] In this embodiment, the first target extrinsic parameter initial value and the second target extrinsic parameter initial value can be understood as optimized first extrinsic parameter initial value and second extrinsic parameter initial value, and the first extrinsic parameter initial value and the second extrinsic parameter initial value can be understood as two initial values representing the extrinsic parameter of the third camera. The third target extrinsic parameter initial value and the fourth target extrinsic parameter initial value can be understood as optimized third extrinsic parameter initial value and fourth extrinsic parameter initial value, and the third extrinsic parameter initial value and the fourth extrinsic parameter initial value can be understood as two initial values representing the extrinsic parameter of the fourth camera.

[0071] Here, how to determine each target extrinsic parameter initial value is not specifically limited; for example, first, an image processing can be performed on the images collected by the cameras through a corresponding algorithm to extract corresponding feature points of the first vehicle body coordinate point obtained by the first camera and the third camera through feature point matching on the third camera image, corresponding feature points of the second vehicle body coordinate point obtained by the first camera and the fourth camera through feature point matching on the fourth camera image, corresponding feature points of the third vehicle body coordinate point obtained by the second camera and the third camera through feature point matching on the third camera image, and corresponding feature points of the fourth vehicle body coordinate point obtained by the second camera and the fourth camera through feature point matching on the fourth camera image. The feature points can represent pixel points with significant features. Then, the initial extrinsic parameter of the third camera and the initial extrinsic parameter of the fourth camera are determined through a corresponding algorithm according to the extracted feature points and the determined vehicle body coordinate points. Finally, the first target extrinsic parameter initial value and the second target extrinsic parameter initial value of the third camera and the third target extrinsic parameter initial value and the fourth target extrinsic parameter initial value of the fourth camera are obtained by optimizing the initial extrinsic parameter of the third camera and the initial extrinsic parameter of the fourth camera through a pre-set optimization algorithm.

[0072] S280, determining the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value according to the first target extrinsic initial value, the second target extrinsic initial value, the third target extrinsic initial value, the fourth target extrinsic initial value, the extrinsic of the first camera and the extrinsic of the second camera.

[0073] In the embodiment, the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value can be obtained through corresponding algorithms according to the first target extrinsic initial value, the second target extrinsic initial value, the third target extrinsic initial value, the fourth target extrinsic initial value, the extrinsic of the first camera and the extrinsic of the second camera.

[0074] S290, determining the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value according to the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value.

[0075] S2100, determining the extrinsic of the third camera and the extrinsic of the fourth camera according to the extrinsic of the first camera, the extrinsic of the second camera, the first target pose initial value and the second target pose initial value, the third target pose initial value and the fourth target pose initial value.

[0076] The embodiment two of the application provides a kind of extrinsic calibration method of vehicle-mounted surround view camera, and the process of determining the extrinsic of the first camera and the extrinsic of the second camera according to the vehicle body odometer information of vehicle, the first vision odometer information of the first camera and the second vision odometer information of the second camera, the process of determining the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value according to vehicle body odometer information, the first vision odometer information, the second vision odometer information, the extrinsic of the first camera and the extrinsic of the second camera are specifically embodied.Utilizing the method, the extrinsic of other two cameras can be determined by the extrinsic of the first two cameras, avoid the problem of large computing power and long time consumption caused by the operation of camera vision odometer of four extrinsics, effectively improve the efficiency of extrinsic calibration, guarantee the real-time performance of extrinsic calibration.

[0077] Optionally, the first effective map point, the second effective map point, the third effective map point and the fourth effective map point are determined according to the vehicle body odometer information, the first map point in the first vision odometer information, the second map point in the second vision odometer information and the collection range of each camera, and the method comprises the following steps.

[0078] The scale factor is determined based on the vehicle body motion trajectory information in the vehicle body odometer information and the camera motion trajectory information in the determined vision odometer information, and the scale factor is a parameter representing the proportion between the scale of real world and the scale of three-dimensional map constructed by the camera.

[0079] scaling the first map point and the second map point based on the scale factor to obtain a first real map point of the first map point in a real world scale and a second real map point of the second map point in the real world scale;

[0080] determining the first overlap region, the second overlap region, the third overlap region and the fourth overlap region according to the capture ranges of the cameras, the first overlap region being an overlap region of the capture ranges between the first camera and the third camera, the second overlap region being an overlap region of the capture ranges between the first camera and the fourth camera, the third overlap region being an overlap region of the capture ranges between the second camera and the third camera, and the fourth overlap region being an overlap region of the capture ranges between the second camera and the fourth camera;

[0081] determining a map point of the first real map point in the first overlap region as a first effective map point, determining a map point of the first real map point in the second overlap region as a second effective map point, determining a map point of the second real map point in the third overlap region as a third effective map point, and determining a map point of the second real map point in the fourth overlap region as a fourth effective map point.

[0082] In the embodiment, the real world scale can be understood as a scale in a real world. The scale of the three-dimensional map constructed by the cameras can be understood as a scale of constructing the three-dimensional map. The vehicle body motion trajectory information can be understood as information representing a vehicle body motion trajectory. The camera motion trajectory information can be understood as information representing a camera motion trajectory. The vehicle can obtain the corresponding scale factor through a corresponding algorithm based on the vehicle body motion trajectory information in the vehicle body odometry information and the camera motion trajectory information in the determined visual odometry information.

[0083] Optionally, the first target extrinsic parameter initial value and the second target extrinsic parameter initial value of the third camera and the third target extrinsic parameter initial value and the fourth target extrinsic parameter initial value of the fourth camera are determined according to the images captured by the cameras, the first vehicle body coordinate point, the second vehicle body coordinate point, the third vehicle body coordinate point and the fourth vehicle body coordinate point, and the determination includes:

[0084] The first target feature point, the second target feature point, the third target feature point and the fourth target feature point are determined according to images collected by the cameras, the first target feature point is a corresponding feature point of a first vehicle body coordinate point obtained by the first camera and the third camera through feature point matching on a third camera image, the second target feature point is a corresponding feature point of a second vehicle body coordinate point obtained by the first camera and the fourth camera through feature point matching on a fourth camera image, the third target feature point is a corresponding feature point of a third vehicle body coordinate point obtained by the second camera and the third camera through feature point matching on the third camera image, and the fourth target feature point is a corresponding feature point of a fourth vehicle body coordinate point obtained by the second camera and the fourth camera through feature point matching on the fourth camera image.

[0085] In the embodiment, how to determine the first target feature point, the second target feature point, the third target feature point and the fourth target feature point according to images collected by the cameras is not specifically limited; the feature points of the first camera represent two-dimensional feature points corresponding to map points in the first visual odometry information. The feature points of the second camera represent two-dimensional feature points corresponding to map points in the second visual odometry information. The first target feature point and the second target feature point are determined by matching the feature points of the first camera with the feature points of the third camera and the fourth camera through a feature matching algorithm. The third target feature point and the fourth target feature point are determined by matching the feature points of the second camera with the feature points of the third camera and the fourth camera through the feature matching algorithm. The feature matching algorithm is not specifically limited here.

[0086] The first extrinsic parameter initial value and the second extrinsic parameter initial value of the third camera and the third extrinsic parameter initial value and the fourth extrinsic parameter initial value of the fourth camera are determined through a Perspective-n-Point (PNP) method according to the vehicle body coordinate points and the target feature points.

[0087] In the embodiment, the extrinsic parameter initial value can be understood as an initial value of an extrinsic parameter. How to determine the extrinsic parameter initial value of the third camera and the extrinsic parameter initial value of the fourth camera according to the vehicle body coordinate points and the target feature points is not specifically limited here; for example, the first extrinsic parameter initial value of the third camera can be obtained based on the first vehicle body coordinate point and the first target feature point through a corresponding algorithm (such as a PNP algorithm), the second extrinsic parameter initial value of the third camera can be obtained based on the second vehicle body coordinate point and the second target feature point, the first extrinsic parameter initial value of the fourth camera can be obtained based on the third vehicle body coordinate point and the third target feature point, and the second extrinsic parameter initial value of the fourth camera can be obtained based on the fourth vehicle body coordinate point and the fourth target feature point.

[0088] For each of the third camera and the fourth camera, for each value within the set region of the camera, the extrinsic initial value of the camera is updated to the value, and a corresponding two-dimensional coordinate point is obtained by projecting the vehicle body coordinate point corresponding to the camera to the image coordinate system of the camera, the photometric error between the two-dimensional coordinate point corresponding to the camera and the target feature point is determined, and the set region is associated with the extrinsic initial value of the corresponding camera;

[0089] In the embodiment, the set region is associated with the extrinsic initial value of the corresponding camera; the set region is not limited here, and can be a set range region determined with the extrinsic initial value as a middle value, or a set range region determined with the extrinsic initial value as a left interval endpoint, etc. The image coordinate system can be understood as a coordinate system with the center of the image plane as a coordinate origin, and X-axis and Y-axis parallel to two vertical edges of the image plane; the image coordinate system can be considered as a coordinate system representing the position of a pixel in the image in physical units (for example, millimeters). The photometric error can be understood as the difference between the photometry of the two-dimensional coordinate point and the photometry of the target feature point. Each value within the set region can correspond to a photometric error.

[0090] The value corresponding to the minimum photometric error in the photometric error corresponding to the camera is determined as the target extrinsic initial value corresponding to the camera.

[0091] In the embodiment, the minimum photometric error can be understood as the photometric error with the minimum value.

[0092] Optionally, the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value are determined according to the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value, including:

[0093] The first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value are determined as the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value.

[0094] Optionally, the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value are determined according to the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value, including:

[0095] If the first photometric error value under the first pose initial value, the second photometric error value under the second pose initial value, the third photometric error value under the third pose initial value, and the fourth photometric error value under the fourth pose initial value do not satisfy the first set condition, each pose initial value is updated based on a first set step size until each photometric error value under the updated pose initial value satisfies the first set condition, and each pose initial value is determined as each target pose initial value.

[0096] The first photometric error value is an error value between a photometric value of the first projection point and a photometric value of the first target feature point, and the first projection point is a point obtained by projecting the first effective map point of the first camera to the image coordinate system of the third camera under the first pose initial value.

[0097] The second photometric error value is an error value between a photometric value of the second projection point and a photometric value of the second target feature point, and the second projection point is a point obtained by projecting the second effective map point of the first camera to the image coordinate system of the fourth camera under the second pose initial value.

[0098] The third photometric error value is an error value between a photometric value of the third projection point and a photometric value of the third target feature point, and the third projection point is a point obtained by projecting the third effective map point of the second camera to the image coordinate system of the third camera under the third pose initial value.

[0099] The fourth photometric error value is an error value between a photometric value of the fourth projection point and a photometric value of the fourth target feature point, and the fourth projection point is a point obtained by projecting the fourth effective map point of the second camera to the image coordinate system of the fourth camera under the fourth pose initial value.

[0100] The first set condition is that the first photometric error value reaches a first threshold value, the second photometric error value reaches a second threshold value, the third photometric error value reaches a third threshold value, and the fourth photometric error value reaches a fourth threshold value.

[0101] In the embodiment, the vehicle can obtain the first pose initial value by setting a pose algorithm according to the extrinsic parameter of the first camera and the target extrinsic parameter initial value of the third camera, and obtain the second pose initial value by setting a pose algorithm according to the extrinsic parameter of the first camera and the target extrinsic parameter initial value of the fourth camera. Correspondingly, the vehicle can obtain the third pose initial value by setting a pose algorithm according to the extrinsic parameter of the second camera and the target extrinsic parameter initial value of the third camera, and obtain the fourth pose initial value by setting a pose algorithm according to the extrinsic parameter of the second camera and the target extrinsic parameter initial value of the fourth camera.

[0102] In this embodiment, the first threshold value can be understood as a first preset threshold value, the second threshold value can be understood as a second preset threshold value, the third threshold value can be understood as a third preset threshold value, and the fourth threshold value can be understood as a fourth preset threshold value. The first threshold value, the second threshold value, the third threshold value, and the fourth threshold value are not limited here. The first luminosity error value reaching the first threshold value can be considered as the first luminosity error value being less than or equal to the first threshold value; correspondingly, the second luminosity error value reaching the second threshold value can be considered as the second luminosity error value being less than or equal to the second threshold value; the third luminosity error value reaching the third threshold value can be considered as the third luminosity error value being less than or equal to the third threshold value; and the fourth luminosity error value reaching the fourth threshold value can be considered as the fourth luminosity error value being less than or equal to the fourth threshold value.

[0103] The first set step can be understood as a preset step, which is not limited here. How to update the first pose initial value, the second pose initial value, the third pose initial value, and / or the fourth pose initial value based on the first set step is not limited here; for example, the first pose initial value, the second pose initial value, the third pose initial value, and / or the fourth pose initial value can be increased by the first set step, or the first pose initial value, the second pose initial value, the third pose initial value, and / or the fourth pose initial value can be decreased by the first set step, etc. It can be understood that after the first pose initial value is updated, the relative pose between the first camera and the third camera will change, and the corresponding first luminosity error will change; correspondingly, after the second pose initial value is updated, the relative pose between the first camera and the fourth camera will change, and the corresponding second luminosity error will change, and so on.

[0104] Optionally, after the extrinsic parameters of the third camera and the extrinsic parameters of the fourth camera are determined, the method further includes:

[0105] Under the extrinsic parameters of the cameras, the error index is determined according to the following formula:

[0106] E rr = E1 + E2 + E3 + E4 + E5 + E6;

[0107] wherein, E rrError indicators are parameter indicators for measuring the optimization of the extrinsic parameters of each camera; E1 is a re-projection error value of the first effective map point of the first camera projected to the image coordinate system of the third camera; E2 is a re-projection error value of the third effective map point of the second camera projected to the image coordinate system of the third camera; E3 is a re-projection error value of the second effective map point of the first camera projected to the image coordinate system of the fourth camera; E4 is a re-projection error value of the fourth effective map point of the second camera projected to the image coordinate system of the fourth camera; E5 is a hand-eye calibration error value determined based on the vehicle body odometer and the first visual odometer of the first camera; E6 is a hand-eye calibration error value determined based on the vehicle body odometer and the second visual odometer of the second camera;

[0108] In the embodiment, the re-projection error can be understood as the pixel difference between the effective map point in the camera coordinate system and the real point in the image coordinate system. The hand-eye calibration error value can be understood as the odometer error value between the vehicle body odometer and the visual odometer calculated based on the hand-eye calibration principle.

[0109] In an embodiment, the optimization target is: arg min ξ (E rr ), wherein, is the Lie algebra representation of Similarly, that is is the Lie algebra representation of is the Lie algebra representation of is the Lie algebra representation of is the Lie algebra representation of

[0110] In an embodiment,

[0111]

[0112]

[0113]

[0114]

[0115]

[0116] wherein, ​​​respectively represent the front, rear, left and right view camera extrinsic parameters, wherein the first camera, the second camera, the third camera and the fourth camera are taken as examples for the front, rear, right and left view cameras respectively for illustrative purposes;

[0117]

[0118] I1(), I2() respectively represent the gray value (or luminosity) of the image feature points;

[0119] A l , A r respectively represent the intrinsic matrix of the left and right view cameras;

[0120] d() represents the distance measure on the Euclidean algebra;

[0121] P f,i ,P re,i respectively represent the front and rear view camera three-dimensional map points;

[0122] u f,i ,u re,i respectively represent the front view camera image feature points and the rear view camera image feature points;

[0123] represents the pose relationship of adjacent image frames in the vehicle body coordinate system;

[0124] represents the pose relationship of adjacent image frames in the camera coordinate system;

[0125] w fl , w rl , w fr , w rr , W 1,2,3,4,5,6 represents the pre-set weight.

[0126] If the error index reaches the error threshold, the extrinsic parameters of each camera are determined as the target extrinsic parameters of each camera, and the target extrinsic parameters are the optimized extrinsic parameters;

[0127] If the error index does not reach the error threshold, the extrinsic parameters of at least one camera are updated based on the second set step size until the error index corresponding to each camera reaches the error threshold, and the extrinsic parameters corresponding to each camera when reaching the error threshold are determined as the target extrinsic parameters of each camera.

[0128] In this embodiment, the error threshold can be understood as a pre-set threshold for measuring the error index, which is not specifically limited here. The second set step size can be understood as a pre-set second step size, which is not specifically limited here.

[0129] How to update the external parameters of the at least one camera based on the second set step is not specifically limited here, such as increasing or decreasing the external parameters of the at least one camera by the second set step to obtain new external parameters.

[0130] After calculating the initial values of the external parameters of the four cameras, the embodiment further improves the external parameter calibration accuracy by jointly optimizing the initial values of the external parameters of the four cameras.

[0131] Optionally, after determining the target external parameters of each camera, the method further comprises:

[0132] For each camera, the Euler angle information and the translation vector information of the camera are determined based on the target external parameters of the camera, the Euler angle information being information representing a rotation angle of the camera relative to the vehicle body coordinate system, and the translation vector information being information representing a translation amount of the camera relative to the vehicle body coordinate system;

[0133] It is determined whether the Euler angle information and the translation vector information of the camera satisfy a second set condition, the second set condition being that the Euler angle information is within a first threshold range and the translation vector information is within a second threshold range;

[0134] If yes, it is determined that the target external parameter calibration of the camera is successful;

[0135] Otherwise, it is determined that the target external parameter calibration of the camera fails.

[0136] In the embodiment, the Euler angle information can include the pitch angle α, the yaw angle β and the roll angle σ of the camera. How to determine the Euler angle information and the translation vector information of the camera based on the target external parameters of the camera is not specifically limited here, such as recognizing and processing the target external parameters of the camera by a corresponding algorithm to obtain the Euler angle information and the translation vector information of the camera.

[0137] The first threshold range can be understood as a first threshold range set in advance; the second threshold range can be understood as a second threshold range set in advance; the first threshold range and the second threshold range are not specifically limited here; such as the first threshold range can be set to 0°<α<90°, -90°<β<90°, -90°<σ<90°.

[0138] Embodiment three

[0139] Figure 5 A structural schematic diagram of a vehicle-mounted surround-view camera external parameter calibration device provided by the third embodiment of the application, which can be implemented by software and / or hardware. For example, Figure 5As shown, the device is configured in a vehicle, the vehicle comprising a vehicle-mounted surround view camera, the vehicle-mounted surround view camera comprising a first camera, a second camera, a third camera and a fourth camera, the first camera position and the third camera position, the fourth camera position being adjacent, the second camera position and the third camera position, the fourth camera position being adjacent, the device comprising:

[0140] A first extrinsic parameter determination module 310 is configured to determine the extrinsic parameters of the first camera and the extrinsic parameters of the second camera according to the vehicle body odometry information of the vehicle, the first visual odometry information of the first camera and the second visual odometry information of the second camera.

[0141] A pose determination module 320 is configured to determine the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value according to the vehicle body odometry information, the first visual odometry information, the second visual odometry information, the extrinsic parameters of the first camera and the extrinsic parameters of the second camera.

[0142] An initial value determination module 330 is configured to determine the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value according to the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value.

[0143] A second extrinsic parameter determination module 340 is configured to determine the extrinsic parameters of the third camera and the extrinsic parameters of the fourth camera according to the extrinsic parameters of the first camera, the extrinsic parameters of the second camera, the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value.

[0144] In this embodiment, the first external parameter determination module determines the external parameters of the first camera and the external parameters of the second camera according to the vehicle body odometer information, the first visual odometer information of the first camera, and the second visual odometer information of the second camera; the pose determination module determines the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value according to the vehicle body odometer information, the first visual odometer information, the second visual odometer information, the external parameters of the first camera, and the external parameters of the second camera; the initial value determination module determines the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value according to the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value; and the second external parameter determination module determines the external parameters of the third camera and the external parameters of the fourth camera according to the external parameters of the first camera, the external parameters of the second camera, the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value. The technical solution can avoid the problems of large computing power and long time consumption caused by the visual odometer calculation of the four cameras to determine the external parameters, thereby improving the real-time performance of the external parameter calibration. In addition, the technical solution can be applied to external parameter calibration in any scene, can avoid the problem of relying on specific road markers, and improves the universality of the external parameter calibration.

[0145] Optionally, the first external parameter determination module 310 comprises:

[0146] The image acquisition unit is configured to acquire the vehicle body odometer information and images captured by the cameras.

[0147] The odometer determination unit is configured to determine the first visual odometer information and the second visual odometer information based on the images captured by the first camera and the second camera.

[0148] The external parameter determination unit is configured to determine the external parameters of the first camera and the external parameters of the second camera by setting an extrinsic-intrinsic calibration algorithm based on the vehicle body odometer information and the determined visual odometer information.

[0149] Optionally, the pose determination module 320 comprises:

[0150] The map point determination unit is configured to determine the first effective map point, the second effective map point, the third effective map point, and the fourth effective map point according to the vehicle body odometer information, the first map point in the first visual odometer information, the second map point in the second visual odometer information, and the capture ranges of the cameras.

[0151] a first conversion unit, configured to convert the first effective map point and the second effective map point into a first coordinate point and a second coordinate point in a first camera coordinate system according to poses of the first visual odometry information and second visual odometry information, and convert the third effective map point and the fourth effective map point into a third coordinate point and a fourth coordinate point in a second camera coordinate system;

[0152] a second conversion unit, configured to convert the first coordinate point, the second coordinate point, the third coordinate point and the fourth coordinate point into a first vehicle body coordinate point, a second vehicle body coordinate point, a third vehicle body coordinate point and a fourth vehicle body coordinate point in a vehicle body coordinate system respectively according to extrinsic parameters of the first camera and extrinsic parameters of the second camera;

[0153] an extrinsic parameter initial value determination unit, configured to determine a first target extrinsic parameter initial value and a second target extrinsic parameter initial value of the third camera, and a third target extrinsic parameter initial value and a fourth target extrinsic parameter initial value of the fourth camera according to images collected by the cameras, the first vehicle body coordinate point, the second vehicle body coordinate point, the third vehicle body coordinate point and the fourth vehicle body coordinate point;

[0154] a pose initial value determination unit, configured to determine a first pose initial value, a second pose initial value, a third pose initial value and a fourth pose initial value according to the first target extrinsic parameter initial value, the second target extrinsic parameter initial value, the third target extrinsic parameter initial value, the fourth target extrinsic parameter initial value, the extrinsic parameters of the first camera and the extrinsic parameters of the second camera.

[0155] Optionally, the map point determination unit comprises:

[0156] a factor determination sub-unit, configured to determine a scale factor based on vehicle body motion trajectory information in the vehicle body odometry information and determined camera motion trajectory information in the visual odometry information, the scale factor being a parameter representing a proportion between a scale of a real world and a scale of a three-dimensional map constructed by a camera;

[0157] a conversion sub-unit, configured to perform scale conversion on the first map point and the second map point based on the scale factor, to obtain a first real map point of the first map point in a real world scale and a second real map point of the second map point in a real world scale;

[0158] a region determining unit configured to determine a first overlap region, a second overlap region, a third overlap region and a fourth overlap region according to the capture ranges of the cameras, the first overlap region being an overlap region of the capture ranges between the first camera and the third camera, the second overlap region being an overlap region of the capture ranges between the first camera and the fourth camera, the third overlap region being an overlap region of the capture ranges between the second camera and the third camera, and the fourth overlap region being an overlap region of the capture ranges between the second camera and the fourth camera;

[0159] a map point determining subunit configured to determine a map point of the first real map point in the first overlap region as a first effective map point, determine a map point of the first real map point in the second overlap region as a second effective map point, determine a map point of the second real map point in the third overlap region as a third effective map point, and determine a map point of the second real map point in the fourth overlap region as a fourth effective map point.

[0160] Optionally, the initial value of the external parameter determining unit comprises:

[0161] a feature point determining subunit configured to determine a first target feature point, a second target feature point, a third target feature point and a fourth target feature point according to the images captured by the cameras, the first target feature point being a corresponding feature point of a first vehicle body coordinate point in the third camera image obtained by feature point matching between the first camera and the third camera, the second target feature point being a corresponding feature point of a second vehicle body coordinate point in the fourth camera image obtained by feature point matching between the first camera and the fourth camera, the third target feature point being a corresponding feature point of a third vehicle body coordinate point in the third camera image obtained by feature point matching between the second camera and the third camera, and the fourth target feature point being a corresponding feature point of a fourth vehicle body coordinate point in the fourth camera image obtained by feature point matching between the second camera and the fourth camera.

[0162] a first initial value determining subunit configured to determine the first external parameter initial value and the second external parameter initial value of the third camera and the third external parameter initial value and the fourth external parameter initial value of the fourth camera by PNP method according to the vehicle body coordinate points and the target feature points.

[0163] an error determining subunit configured to, for each of the third camera and the fourth camera, update the external parameter initial value of the camera to each value in a set region of the camera, project the vehicle body coordinate point corresponding to the camera into the image coordinate system of the camera to obtain a corresponding two-dimensional coordinate point, and determine the photometric error between the corresponding two-dimensional coordinate point and the target feature point of the camera, the set region being associated with the external parameter initial value of the corresponding camera.

[0164] The second initial value determination subunit is configured to determine a value corresponding to a minimum photometric error in the photometric errors corresponding to the cameras as the initial value of the target extrinsic parameter corresponding to the cameras.

[0165] Optionally, the initial value determination module 330 comprises:

[0166] The first initial value determination unit is configured to determine the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value as the first target pose initial value, the second target pose initial value, the third target pose initial value and the fourth target pose initial value.

[0167] Optionally, the initial value determination module 330 further comprises:

[0168] The second initial value determination unit is configured to, if the first photometric error value under the first pose initial value, the second photometric error value under the second pose initial value, the third photometric error value under the third pose initial value and the fourth photometric error value under the fourth pose initial value do not satisfy the first set condition, update the pose initial values based on a first set step size until the photometric error values under the updated pose initial values satisfy the first set condition, and determine the pose initial values as the target pose initial values.

[0169] The first photometric error value is an error value between the photometric value of the first projection point and the photometric value of the first target feature point, and the first projection point is a point obtained by projecting the first valid map point of the first camera to the image coordinate system of the third camera under the first pose initial value.

[0170] The second photometric error value is an error value between the photometric value of the second projection point and the photometric value of the second target feature point, and the second projection point is a point obtained by projecting the second valid map point of the first camera to the image coordinate system of the fourth camera under the second pose initial value.

[0171] The third photometric error value is an error value between the photometric value of the third projection point and the photometric value of the third target feature point, and the third projection point is a point obtained by projecting the third valid map point of the second camera to the image coordinate system of the third camera under the third pose initial value.

[0172] The fourth photometric error value is an error value between the photometric value of the fourth projection point and the photometric value of the fourth target feature point, and the fourth projection point is a point obtained by projecting the fourth valid map point of the second camera to the image coordinate system of the fourth camera under the fourth pose initial value.

[0173] The first set condition is that the first luminosity error value reaches a first threshold value, the second luminosity error value reaches a second threshold value, the third luminosity error value reaches a third threshold value, and the fourth luminosity error value reaches a fourth threshold value.

[0174] Optionally, the apparatus further comprises:

[0175] An index determination module is configured to, after determining the extrinsic parameters of the third camera and the extrinsic parameters of the fourth camera, determine an error index according to the following formula under the extrinsic parameters of each camera:

[0176] E rr = E1 + E2 + E3 + E4 + E5 + E6.

[0177] Wherein, E rr is the error index, which is a parameter index for measuring the optimization of the extrinsic parameters of each camera; E1 is a re-projection error value of the first effective map point of the first camera projected to the image coordinate system of the third camera; E2 is a re-projection error value of the third effective map point of the second camera projected to the image coordinate system of the third camera; E3 is a re-projection error value of the second effective map point of the first camera projected to the image coordinate system of the fourth camera; E4 is a re-projection error value of the fourth effective map point of the second camera projected to the image coordinate system of the fourth camera; E5 is a hand-eye calibration error value determined based on the vehicle body odometer and the first visual odometer of the first camera; E6 is a hand-eye calibration error value determined based on the vehicle body odometer and the second visual odometer of the second camera;

[0178] A third extrinsic parameter determination module is configured to, if the error index reaches an error threshold value, determine the extrinsic parameters of each camera as target extrinsic parameters of each camera, the target extrinsic parameters being optimized extrinsic parameters.

[0179] A fourth extrinsic parameter determination module is configured to, if the error index does not reach the error threshold value, update the extrinsic parameters of at least one camera based on a second set step size until the error index corresponding to each camera reaches the error threshold value, and determine the extrinsic parameters of each camera corresponding to the error threshold value as target extrinsic parameters of each camera.

[0180] Optionally, the apparatus further comprises:

[0181] An information determination module is configured to, after determining the target extrinsic parameters of each camera, determine, for each camera, Euler angle information and translation vector information of the camera based on the target extrinsic parameters of the camera, the Euler angle information being information representing a rotation angle of the camera relative to a vehicle body coordinate system, and the translation vector information being information representing a translation amount of the camera relative to the vehicle body coordinate system.

[0182] a judging module configured to judge whether the Euler angle information and the translation vector information of the camera satisfy a second set condition, the second set condition being that the Euler angle information is within a first threshold range and the translation vector information is within a second threshold range;

[0183] a first calibration module configured to, if yes, determine that target extrinsic parameter calibration of the camera is successful;

[0184] a second calibration module configured to, if no, determine that target extrinsic parameter calibration of the camera fails.

[0185] The vehicle-mounted surround-view camera extrinsic parameter calibration device provided in the embodiments of the present application can execute the vehicle-mounted surround-view camera extrinsic parameter calibration method provided in any of the embodiments of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0186] Embodiment Four

[0187] Figure 6 A structural schematic diagram of a vehicle is provided in Embodiment Four of the present application. Figure 6 As shown in the figure, the vehicle 10 includes a vehicle-mounted surround-view camera 20, at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is in communication connection with the at least one processor 11, wherein the memory stores a computer program that can be executed by the at least one processor, and the processor 11 can execute various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the vehicle 10 can also be stored. The vehicle-mounted surround-view camera 20, the processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0188] A plurality of components in the vehicle 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, a loudspeaker, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the vehicle 10 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.

[0189] The processor 11 can be various general and / or special purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, and the like. The processor 11 performs various methods and processes described above, such as the extrinsic calibration method of the vehicle surround view camera.

[0190] In some embodiments, the extrinsic calibration method of the vehicle surround view camera can be implemented as a computer program tangibly embodied in a computer readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the vehicle 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded onto the RAM 13 and executed by the processor 11, one or more steps of the extrinsic calibration method of the vehicle surround view camera described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the extrinsic calibration method of the vehicle surround view camera by any other suitable means, such as by means of firmware.

[0191] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (PLD), a computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0192] Computer programs used to implement the methods of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor of the machine, implements the functions / acts specified in the flowcharts and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0193] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer-readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of a program of instructions in a transitory signal, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0194] To provide for interaction with a user, the systems and techniques described here can be implemented on a vehicle having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the vehicle. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0195] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0196] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0197] It should be understood that the various forms of flow shown above can be reordered, added to, or have steps deleted. For example, the steps described in the present application can be performed in parallel, in series, or in a different order, as long as the desired results of the technical solutions of the present application can be achieved, which are not limited herein.

[0198] The above detailed description does not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for calibrating the extrinsic parameters of a vehicle-mounted surround-view camera, characterized in that, Applied to a vehicle, the vehicle includes an onboard surround-view camera, the onboard surround-view camera including a first camera, a second camera, a third camera, and a fourth camera, wherein the positions of the first camera, the third camera, and the fourth camera are adjacent, and the positions of the second camera, the third camera, and the fourth camera are adjacent, the method includes: Based on the vehicle's odometer information, the first visual odometer information of the first camera, and the second visual odometer information of the second camera, the extrinsic parameters of the first camera and the second camera are determined. Based on the vehicle odometer information, the first visual odometer information, the second visual odometer information, the extrinsic parameters of the first camera and the extrinsic parameters of the second camera, determine the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value; The initial values ​​of the first target pose, the second target pose, the third target pose, and the fourth target pose are determined based on the first initial pose value, the second initial pose value, the third initial pose value, and the fourth initial pose value. The extrinsic parameters of the third camera and the fourth camera are determined based on the extrinsic parameters of the first camera, the extrinsic parameters of the second camera, the initial pose values ​​of the first target, the second target, the third target, and the fourth target.

2. The method according to claim 1, characterized in that, Based on the vehicle odometer information, the first visual odometer information, the second visual odometer information, the extrinsic parameters of the first camera, and the extrinsic parameters of the second camera, the initial values ​​of the first pose, the second pose, the third pose, and the fourth pose are determined, including: Based on the vehicle odometer information, the first map point in the first visual odometer information, the second map point in the second visual odometer information, and the acquisition range of each camera, the first effective map point, the second effective map point, the third effective map point, and the fourth effective map point are determined. Based on the pose of the first visual odometry information and the second visual odometry information, the first effective map point and the second effective map point are transformed into the first camera coordinate system to obtain the first coordinate point and the second coordinate point, and the third effective map point and the fourth effective map point are transformed into the second camera coordinate system to obtain the third coordinate point and the fourth coordinate point. Based on the extrinsic parameters of the first camera and the second camera, the first coordinate point, the second coordinate point, the third coordinate point and the fourth coordinate point are transformed into the vehicle coordinate system to obtain the first vehicle coordinate point, the second vehicle coordinate point, the third vehicle coordinate point and the fourth vehicle coordinate point; Based on the images captured by each camera, the first vehicle body coordinate point, the second vehicle body coordinate point, the third vehicle body coordinate point, and the fourth vehicle body coordinate point, determine the initial values ​​of the first and second target extrinsic parameters of the third camera, as well as the initial values ​​of the third and fourth target extrinsic parameters of the fourth camera. Based on the initial values ​​of the first target extrinsic parameters, the second target extrinsic parameters, the third target extrinsic parameters, the fourth target extrinsic parameters, the extrinsic parameters of the first camera, and the extrinsic parameters of the second camera, the initial values ​​of the first pose, the second pose, the third pose, and the fourth pose are determined.

3. The method according to claim 1, characterized in that, Determining the initial values ​​of the first target pose, the second target pose, the third target pose, and the fourth target pose based on the first initial pose value, the second initial pose value, the third initial pose value, and the fourth initial pose value includes: The first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value are determined as the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value.

4. The method according to claim 2, characterized in that, Determining the initial values ​​of the first target pose, the second target pose, the third target pose, and the fourth target pose based on the first initial pose value, the second initial pose value, the third initial pose value, and the fourth initial pose value includes: If the first photometric error value under the first pose initial value, the second photometric error value under the second pose initial value, the third photometric error value under the third pose initial value, and the fourth photometric error value under the fourth pose initial value do not meet the first set condition, then each of the pose initial values ​​is updated based on the first set step size until each of the photometric error values ​​under the updated pose initial values ​​meets the first set condition, and each of the pose initial values ​​is determined as the target pose initial value. Wherein, the first photometric error value is the error value between the photometric value of the first projection point and the photometric value of the first target feature point, and the first projection point is the point obtained by projecting the first effective image point of the first camera onto the image coordinate system of the third camera under the first pose initial value. The second photometric error value is the error between the photometric value of the second projection point and the photometric value of the second target feature point. The second projection point is the point obtained by projecting the second effective image point of the first camera onto the image coordinate system of the fourth camera under the second pose initial value. The third photometric error value is the error value between the photometric value of the third projection point and the photometric value of the third target feature point. The third projection point is the point obtained by projecting the third effective image point of the second camera onto the image coordinate system of the third camera under the third pose initial value. The fourth photometric error value is the error value between the photometric value of the fourth projection point and the photometric value of the fourth target feature point. The fourth projection point is the point obtained by projecting the fourth effective mapping point of the second camera onto the image coordinate system of the fourth camera under the fourth pose initial value. The first set condition is that the first photometric error value reaches the first threshold, the second photometric error value reaches the second threshold, the third photometric error value reaches the third threshold, and the fourth photometric error value reaches the fourth threshold.

5. The method according to claim 2, characterized in that, Based on the vehicle odometer information, the first map point in the first visual odometer information, the second map point in the second visual odometer information, and the acquisition range of each camera, a first effective mapping point, a second effective mapping point, a third effective mapping point, and a fourth effective mapping point are determined, including: The scale factor is determined based on the vehicle motion trajectory information in the vehicle odometer information and the camera motion trajectory information in the determined visual odometer information. The scale factor is a parameter that characterizes the ratio between the scale of the real world and the scale of the three-dimensional map constructed by the camera. Based on the scale factor, the first map point and the second map point are scaled to obtain a first real map point of the first map point at a real-world scale and a second real map point of the second map point at a real-world map scale. Based on the acquisition range of each camera, a first overlapping region, a second overlapping region, a third overlapping region, and a fourth overlapping region are determined. The first overlapping region is the overlapping region of the acquisition range between the first camera and the third camera. The second overlapping region is the overlapping region of the acquisition range between the first camera and the fourth camera. The third overlapping region is the overlapping region of the acquisition range between the second camera and the third camera. The fourth overlapping region is the overlapping region of the acquisition range between the second camera and the fourth camera. The map point of the first real map point in the first overlapping area is determined as the first effective map point, the map point of the first real map point in the second overlapping area is determined as the second effective map point, the map point of the second real map point in the third overlapping area is determined as the third effective map point, and the map point of the second real map point in the fourth overlapping area is determined as the fourth effective map point.

6. The method according to claim 2, characterized in that, Based on the images captured by each camera, the first vehicle body coordinate point, the second vehicle body coordinate point, the third vehicle body coordinate point, and the fourth vehicle body coordinate point, determine the initial values ​​of the first and second target extrinsic parameters for the third camera, and the initial values ​​of the third and fourth target extrinsic parameters for the fourth camera, including: Based on the images captured by each camera, a first target feature point, a second target feature point, a third target feature point, and a fourth target feature point are determined. The first target feature point is the corresponding feature point of the first vehicle body coordinate point obtained by feature point matching between the first and third cameras on the third camera image. The second target feature point is the corresponding feature point of the second vehicle body coordinate point obtained by feature point matching between the first and fourth cameras on the fourth camera image. The third target feature point is the corresponding feature point of the third vehicle body coordinate point obtained by feature point matching between the second and third cameras on the third camera image. The fourth target feature point is the corresponding feature point of the fourth vehicle body coordinate point obtained by feature point matching between the second and fourth cameras on the fourth camera image. Based on the vehicle body coordinates and target feature points, the initial values ​​of the first and second extrinsic parameters of the third camera, as well as the initial values ​​of the third and fourth extrinsic parameters of the fourth camera, are determined using the perspective n-point positioning PNP method. For each of the third and fourth cameras, for each value within the set area of ​​the camera, the initial value of the camera's extrinsic parameters is updated to the value, and the vehicle body coordinate point corresponding to the camera is projected onto the image coordinate system of the camera to obtain the corresponding two-dimensional coordinate point. The photometric error between the two-dimensional coordinate point corresponding to the camera and the target feature point is determined, and the set area is associated with the initial value of the corresponding camera's extrinsic parameters. The value corresponding to the smallest photometric error among the photometric errors corresponding to the camera is determined as the initial value of the target extrinsic parameter corresponding to the camera.

7. The method according to claim 1, characterized in that, After determining the extrinsic parameters of the third camera and the fourth camera based on the extrinsic parameters of the first camera, the second camera, the first target pose initial value, the second target pose initial value, the third target pose initial value, and the fourth target pose initial value, the further step includes: Under the various camera extrinsic parameters, the error index is determined according to the following formula: E rr =E1+E2+E3+E4+E5+E6; Among them, E rr The error index is a parameter that measures the optimization of the extrinsic parameters of each camera; E1 is the reprojection error value of the first effective mapping point of the first camera projected onto the image coordinate system of the third camera; E2 is the reprojection error value of the third effective mapping point of the second camera projected onto the image coordinate system of the third camera; E3 is the reprojection error value of the second effective mapping point of the first camera projected onto the image coordinate system of the fourth camera; E4 is the reprojection error value of the fourth effective mapping point of the second camera projected onto the image coordinate system of the fourth camera; E5 is the hand-eye calibration error value determined based on the vehicle odometer and the first visual odometer of the first camera; E6 is the hand-eye calibration error value determined based on the vehicle odometer and the second visual odometer of the second camera. If the error index reaches the error threshold, the extrinsic parameters of each camera are determined as the target extrinsic parameters of each camera, and the target extrinsic parameters are the optimized extrinsic parameters; If the error index does not reach the error threshold, the extrinsic parameters of at least one camera are updated based on the second set step size until the error index corresponding to each camera reaches the error threshold, and the extrinsic parameters of each camera that reach the error threshold are determined as the target extrinsic parameters of each camera.

8. The method according to claim 6, characterized in that, After determining the target extrinsic parameters for each camera, the following is also included: For each camera, the Euler angle information and translation vector information of the camera are determined based on the target extrinsic parameters of the camera. The Euler angle information is information representing the rotation angle of the camera relative to the vehicle coordinate system, and the translation vector information is information representing the translation amount of the camera relative to the vehicle coordinate system. Determine whether the Euler angle information and translation vector information of the camera meet the second set condition, wherein the Euler angle information is within a first threshold range and the translation vector information is within a second threshold range; If so, then the target extrinsic parameter calibration of the camera is confirmed to be successful; Otherwise, the target extrinsic parameter calibration of the camera is determined to have failed.

9. A device for calibrating the extrinsic parameters of a vehicle-mounted surround-view camera, characterized in that, The device is configured in a vehicle, the vehicle including an in-vehicle surround-view camera, the in-vehicle surround-view camera including a first camera, a second camera, a third camera and a fourth camera, the first camera being located adjacent to the third camera and the fourth camera, the second camera being located adjacent to the third camera and the fourth camera, the device comprising: The first extrinsic parameter determination module is used to determine the extrinsic parameters of the first camera and the second camera based on the vehicle's body odometer information, the first visual odometer information of the first camera, and the second visual odometer information of the second camera. The pose determination module is used to determine the first pose initial value, the second pose initial value, the third pose initial value and the fourth pose initial value based on the vehicle odometer information, the first visual odometer information, the second visual odometer information, the extrinsic parameters of the first camera and the extrinsic parameters of the second camera. An initial value determination module is used to determine the initial values ​​of a first target pose, a second target pose, a third target pose, and a fourth target pose based on the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value. The second extrinsic parameter determination module is used to determine the extrinsic parameters of the third camera and the fourth camera based on the extrinsic parameters of the first camera, the extrinsic parameters of the second camera, the initial value of the first target pose, the initial value of the second target pose, the initial value of the third target pose, and the initial value of the fourth target pose.

10. A vehicle, characterized in that, The vehicles include: Vehicle-mounted surround view camera; At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the extrinsic parameter calibration method for the vehicle surround view camera according to any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the extrinsic parameter calibration method for the vehicle surround-view camera as described in any one of claims 1-8.

Images